The gene discovery component of the UNC Breast Spore attempts to identify novel gene products involved in breast cell growth and differentiation. The emphasis is on molecular techniques that have a reasonable chance of isolating new sequences from candidate regulatory proteins. The cDNA sequence can then be quickly used to produce antibodies and probes to scan breast and breast cancer samples for under or overexpression of the putative regulatory protein. We are concentrating on two classes of growth regulators, protein kinases (Ed Liu/Bill Cance) and substrates for the EGF receptor (EGFR) family of tyrosine kinases, particularly HER2 and HER4. The rationale for studying EGFR family member substrates is strong. EGFR and HER2 are overexpressed in many breast cancers; most, but not all of these tumors have a poor prognosis. The less than perfect correlation between receptor expression and prognosis may be due to failure to control for a third variable, e.g. expression of another receptor or the substrates that transmit receptor signals. In the last 2 years, two new members of the EGFR family were cloned (HER3 and HER4) and a putative ligand for HER2 was isolated (NDF/heregulin) that stimulates growth or differentiation depending on the breast cell line used. Very recent data suggest that Heregulin is actually a HER4 (and HER3) ligand that binds HER4 and stimulates physical association with HER2 activating both HER4 and HBR2 tyrosine kinases. Thus Heregulin's action (growth vs. differentiation) may depend on the breast cell's mix of EGFR family members or the expression of post-receptor regulatory proteins. We will use the fact that receptor substrates and downstream adapter proteins contain a region, the SH2 domain, that specifically binds with high affinity to tyrosine phosphorylated sequences in the activated autophosphorylated receptor. Thus, the choice of receptor substrates is determined in part by the sequences in the receptor C-terminus; sequences that are quite divergent in EGFR, HER2, and HER4. While a number of EGF receptor substrates have been identified, specific substrates regulating HER2 and HER4 action have not been discovered even though divergent actions such as differentiation suggest that HER2 and 4 may have at least some specific substrates. We will use 32P-labeled HER2 and HER4 C-terminal sequences to screen breast cancer cDNA expression libraries to isolate SH2 domain encoding cDNAs for HER2 and HER4 substrates. Potential substrate cDNAs and antibodies will be used as probes to analyze expression in breast cancers. Additionally gain or loss of function mutations in substrate genes may be found to predisposed women to breast cancer. Lastly, identification of members of the growth or differentiation signaling pathways may provide novel opportunities for therapy. It is crucial to understand how one ligand can lead to growth (which enhances neoplastic progression) or differentiation (which could slow growth); downstream substrates are a key component.